Metraipolis › archive › dc_ddfd9a6fe968

Uruqi: Learning Spatial Cognition from Visual Experience — arXiv:2609.39195

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.39195
License (as archived)
CC0-1.0
Archived
2026-10-07T14:45:58.783Z
SHA-256 of the archived text
e16234af19a50d2f1434957971a9556a75b3fa95c5c855c07947d20fda24477b
Stable URL of this page
https://metraipolis.com/documento/dc_ddfd9a6fe968

How to cite

Uruqi: Learning Spatial Cognition from Visual Experience — arXiv:2609.39195 — arXiv (en), https://arxiv.org/abs/2609.39195 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ddfd9a6fe968 (Metraipolis archive dc_ddfd9a6fe968, sha256 e16234af19a50d2f1434957971a9556a75b3fa95c5c855c07947d20fda24477b)
BibTeX
@misc{dcddfd9a6fe968,
  title = {Uruqi: Learning Spatial Cognition from Visual Experience — arXiv:2609.39195},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.39195},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ddfd9a6fe968; sha256 e16234af19a50d2f1434957971a9556a75b3fa95c5c855c07947d20fda24477b},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.39195. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Spatial intelligence requires maintaining a coherent understanding of the world as the embodied agent moves. Like humans, the agent must use its own motion to interpret changes across observations and update object locations and spatial relations accordingly. Despite spatial post-training having substantially broadened the spatial intelligence of vision-language models (VLMs), they still struggle with two atomic spatial capabilities: tracking self-motion and mapping the surrounding world during motion. To address this gap, we provide dense multi-turn supervision over interleaved atomic capabilities within each training episode, mimicking the visual experience of a continuously moving agent that reasons as it observes. To scale this up, we synthesize 11,738 motif-driven camera trajectories over a broad range of 3D scenes, supporting self-motion tracking, persistent object mapping, and rich spatial operations within each visual experience. By training models to reason over these atomic questions, our URUQI$_{\mathrm{Syn}}$-8B improves accuracy from 15.84% to 47.73% on our Uruqi benchmark comprising 52k questions across 2.7k episodes. URUQI-SI-Mix-8B further reaches 50.41%, comparable to the 50.08% achieved by GPT-6 Astra. Trained solely on our synthesized data, URUQI$_{\mathrm{Syn}}$-8B achieves an average relative accuracy improvement of 17.13% over its InternVL3-8B backbone across three external spatial benchmarks. These results highlight continuous visual experience as a scalable source of supervision for developing spatial cognition in VLMs.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-10-01 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2609.39195 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-07T14:19:00.707Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-10-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).